★ Climate Change ★ The Smart Grid ★ Universal Basic Income ★ The Automation Crisis ★
UBI is attractive because it appears to meet certain inevitable needs. Automation will eventually create sky-high unemployment rates, thereby triggering social unrest. However, the second and third order effects of UBI’s implementation are catastrophic, so it is incredibly unwise to assume UBI to be an inevitable economic development. It is not the need for UBI that is exigent, but the need to handle unavoidably high rates of unemployment.
I’m not trying to win a popularity contest here.
Why do you hear so much from Silicon Valley entrepreneurs about UBI? Because they recognize the effects their products will unavoidably have on labor markets and are wise enough to position themselves in anticipation. As more products are virtualized and more nodes in the economy are singularized into Silicon Valley’s products, where do the jobs go and what happens as currency flows into the hands of fewer and fewer? So if UBI dispenses tons of artificial liquidity into the economy, who do you think will benefit most as products are virtualized?
What do Silicon Valley apps and all virtualized products monetize? Essentially, your time and attention. This is especially true of social media products. Media should not advocate for UBI because the have they most to gain from consumers’ collective idle time.
Still, Silicon Valley’s desire to posture themselves to handle unemployment is not selfishly motivated. They know it’s real and genuinely want handle this, but the people affected will look for people in power who profited from business and policy.
Computational complexity of specific problems determines the degree to which automation affects the economy, as well as this trend’s timing, pacing and completeness. Additionally, quantum computing’s availability and applicability of causes a second major wave in the trend of automation.
The correct answer is an option for universal service. That is, the government becomes the employer of last resort. This solution avoids product-classification, artificial regulation of markets, the complexity of a possible dual currency system and all the resulting economic distortion. However, this solution does not avoid the need for the government to finance this labor. National service also requires that we establish logistics and operations to mobilize between 20 and 100 million Americans. This must be done five to ten years ahead of time.
The main advantage to National Service over UBI is that what we build ends up building us. We need to rebuild infrastructure anyways.
UBI can be packaged in various forms, including tax credits and other systems of wealth redistribution. The pragmatic implementation of UBI is disappointing and even when expenditures of UBI cash aren’t regulated, it still results in economic distortion and disruption of incentivation structures. In the long-term, it is economically unsustainable, but the problems it addresses are unavoidable.
UBI where expenses are controlled ends up being a transition into a command economy. Several of the following criticisms below only apply to UBI when its usage is regulated, but can still be extended to be criticism of tightly regulated economies in general.
UBI just will not work without mechanisms in place to ensure that people spend UBI on actual living expenses. It requires infrastructure for monitoring purchases and reprimanding abuses of UBI income. Without restricting UBI expenditures, there will be individuals and groups who will siphon away the government liquidity injected into society. Those groups include your friendly neighborhood drug dealer as well as corporations like Starbucks selling lattes for $5.
The system of product classifications essentially creates an economic system with two currencies. Instead of a simple dollar for all debts, public and private, the dollar becomes the currency of the rich and employed, who can buy whatever they want. For the rest of us, an actual UBI policy meant to mitigate unintended consequences instead produces another unintended consequence: a pseudo-currency. The UBI dollar is not for all debts, public and private. It is for all debts that are valid living expenses. This essentially turns the government into your mom, who might give you money as long as it is spent on something she knows you really, really need.
Think about the implications of this in context with the UBI’s raison d’etre. Our country would only ever consider UBI because we need to mitigate the effects of unemployment rates ranging from 15-50%. Without some policy in place to handle the second and third order effects of automation, unemployment rates such as these are realistic for the 2030’s. If UBI is the policy utilized to ameliorate this, then for 15-50% of the population without full-time employment, their income is almost completely in UBI pseudo-dollars. The set of products a UBI-dependent person can purchase is almost totally limited to government-approved products.
If the government restricts the purchasing power of a major segment of population, this means that only UBI-approved products are viable in the marketplace. To increase the marketability of products in a UBI system, then lobbying to have it approved can make or break your business model. This inserts bureaucracy into every facet of life. In a very simple way, it gives the government an unimaginable amount of power and sets up the government to be used as a tool to achieve economic ends. So, what happens?
A massive incentive results to lobby for the reclassification of products and to enact major changes in UBI policy implementation. In a regulated UBI system, what’s worth lobbying for? Increased marketability. Yet, restricting the purchasing power of UBI via product classification is the only way the government can fairly dispense cash to the chronically unemployed without having the liquidity spent on Starbucks lattes and other vices. If the corporate lobbyists win and everything is classified as a UBI product, it doesn’t work. And on the flip side, a dual-currency system where nothing is marketable unless lobbied to regulators Washington places power in the hands of regulators and causes economic turbulence.
Regulatory frameworks can be manipulated via suggestion or outright subversion, amplifying inefficiencies in tightly regulated markets. This leads to a less competitive domestic economy, compared to other players in the global economy. Manipulating regulatory rules and the institutions that create them is a weapon in economic warfare. This weapon becomes more powerful as the degree of amplification that second and third order effects via regulatory rules have on markets. Additionally, the number of regulatory rules needed in a UBI system grows as the system evolves.
Worst of all, even if groups or corporations don’t leverage regulation as a tool for economic manipulation, the government is unable to perform its responsibilities. This leads to inefficient markets, fluctuating prices for both products and financial assets, and general price increases for products across the economy.
Most studies will focus on first or second order effects of UBI, but they will do so in isolation. Even when UBI is provided to entire communities or villages, it is fundamentally different than when provided to an entire region, state or nation. And in this way, the UBI studies will primarily identify positive results and neglect the systemic macroeconomic effects. This will result in cognitive bias, where myopic studies crowd out common sense and dialogue about systemic response to UBI.
The cost of living varies dramatically from place to place. As climate change alters weather patterns, agricultural infrastructure will need to be repositioned. In particular, the price and availability of water will soon have a dramatic affect on the cost of living and the cost of food. Real estate prices will fluctuate and most of this geoeconomic variation over time will be unpredictable. This means we need to invest our resources wisely and be prepared to refinance new infrastructure. Strong dynamics in geoeconomic variation will make it difficult to fairly distribute UBI and to do so such that areas hard-hit by geoeconomic variation will have received a fair sum of money over longer periods of time. The more the sum received deviates from what is fair increases with higher geoeconomic variation over time, which is unpredictable and results from climate change, among other causes.
Wealth redistribution is essentally theft. It is difficult to ethically or morally justify redistribution of wealth, especially since we are incapable of doing so fairly.
Disappointing NEET’s and basement dwellers nationwide, the pragmatic implementation of UBI creates economic turbulence and renders the domestic economy vulnerable to global competition. The regulation stagnates the process by which new products can become marketable. Dismantled incentivization structures stifle innovation. The inescapable economic trends essentiall render each of us into a serf.
This is why I cared so much about launching a startup product from 2012 through 2016. Understanding the inherently finite nature of innovation, I didn’t want to be a serf.
Business is about developing assumptions in contrast to the prevailing understanding that enables someone to posture their product or service in anticipation of changing trends. This aspect of business is fundamental. The shifting long-term objectives and confusing implementation details of regulatory policy, as they are commonly understood by the public, lead to an inability to make lasting assumptions about economic trends. In anticipation of shifting policy that invalidates these assumptions, investors and entrepreneurs evaluating a business model will mostly refrain from investing. There’s no point in financing a business whose model is invalidated by policy changes, potential or inevitable.
Generally, any detraction from the ability to make valid predictive assumptions that hold true over time impairs the identification and execution of economically viable opportunities. Cost and risk become difficult to measure or are measured using a model invalidated by future policy changes. If one can assume that there are few policy changes, then investors have a much larger risk appetite, which leads to higher growth amid stable policy conditions. Heavily regulated environments with high policy turnover, high barrier to entry and an inherently limited market do not promote innovation.
These generalizations about business are related to induction puzzles, where the ability to make assumptions via inductive logic allows one to carve away at the problem space. One uses induction to combinatorially evaluate valid configurations of the system. If more induction is required, then the difficulty of the problem is exponentially more complex. Potential policy changes essentially add variables to be combinated with induction and evaluated for optimization and/or maximization. That’s why you should diversify your asset portfolio.
Capitalism essentially harnesses the same math behind machine learning to create evolutionary dynamics. These “reified maths” form the invisible hand that confers advantage to the best marketed, best formulated or best packaged products. It is quite literally the very same maths behind machine learning. Free market systems reduce inefficiencies in an economy, delivering more value to citizens domestically and making the economy more competitive globally.
A poor implementation of UBI will trigger a collapse of capitalism. Capitalism isn’t capitalism at all when an authority artificially and severely restricts choice. This places enormous power in the hands of regulators in Washington and the lobbyists who court them. The regulators decide who wins and who loses in a market and they are inept at doing so.
The scale of market inefficiency increases by triggering a cascading increase in prices across all types of products, since any product or service has a list of component products in its dependency tree. So, price increases are higher for the consumer because those products are higher up the food chain, so to speak.
This dynamic creates a feedback loop, which is just one of many between the mechanisms for amount/distribution of UBI provided and the amplification of UBI-driven effects in the economy like price stability and market efficiency. The effects are more strongly pronounced when UBI transactions are regulated.
The price inflation above is related to the relative value and demand of goods. It may occur alongside currency inflation, which is all but certain because the tax volume cannot be established through competitive and profitable corporate enterprise, which means government expenditures must be financed through credit via central banks. From a systems theory perspective, the price inflation and currency inflation effects do not add, they multiply. And, given the strong need to finance with credit, the dependency on central banking then strongly influences international currency markets and foreign currency reserves. The question is not whether expenditures and credit of this magnitiude will influence currency markets and reserves, but how.
Economics without mathematics modeled on graph theory is dangerously oversimplified. Linear and non-linear systems modeled on top of graphs can be understood at a high level through analysis of feedback loops and systems theory. Economic manipulation and micromanagement inevitably results in unintended amplification to these feedback loops, which results:
Poorly priced assets
Misguided behavior patterns
Misallocation of wealth
And wholly artificial trends driven by small groups of elitists with the right connections and influence.
Generally speaking, the combinations of positive and negative feedback loops in such a system compose to amplify with magnitude of products of individual effects and not sums. Such systems function healthiest and can be most reliably predicted when left to function for themselves and when artificial manipulation is restricted or at least transparent.
The perceived and actual ROI of employment roles depends on the degree to which each role’s tasks can be automated. For roles involving a more complicated arrangement of tasks, the ability to completely automate that role is inherently diminished. The structuring of roles to task types changes as the labor markets evolve. Generally, if an employment role involves a simpler set of task types, that role can be automated away more quickly. If the set of task types is more complex, there is a greater need for generalized AI. For each task type, the complexity of the solutions nearly completely determines its potential to be automated.
Uber is a perfect example of how automation gradually overtakes employment roles within a business or industry. Uber essentially automated the taxi dispatcher and the task of navigation for drivers.
Once the problem of driving is tractibly solved, it would appear economically advantageous to automate the driver. Uber may decide that it wants to continue offering drivers or some such formulation of its product, since a driver affects the customer experience. Also, with drivers, Uber has a platform composed of people invested in Uber’s success. These drivers also act as vectors for propagating dynamics through social networks (i.e. marketing).
There are roles where automation appears to pay off on the surface, but may not be. In the long term, automation may instead strongly detract from certain aspects of your business and product, especially the customer experience. Total automation removes human social dimensions from your product. Factors like these were ubiquitous before the 21st century, so effects from their presence or absence in a product is not well understood.
It’s a common assumption, but mental labor is not safe. In fact, these are some of the easiest and most effective roles to automate. For roles involving mental labor, a trend will be to automate stages in the role’s decision making process, exposing vital data points in a user interface. From there it’s mostly thumbs up or thumbs down: the employee in the mental labor role makes decisions as data is presented to them. This essentially harnesses as dimensionality reduction, which is required for mitigating computational complexity in machine learning.
With respect to dimensionality reduction, the curse of dimensionality is not kind to humanity in the age of automation, where it before allowed us to solve problems which would otherwise be intractible. For almost any task type, if the complete analytical solution is intractible, there’s usually an efficient formulation involving clustering and dimensionality reduction. This forms the basis of a tractible analytical approximation to the original solution.
Groundhog Day: on February 2nd, to predict the weather for the next six weeks, a computer can either completely process all the differential equations to predict the weather, or it can watch and see whether Punxsutawney Phil returns to his hole.
Of course, there is no actual connection between the two. But without needing to understanding the connection, a computer can identify seemingly uncorrelated variables which can be used to construct an approximate solution. The more data and the more proficient the training, the more accurately its construction of approximated solutions can be leveraged.
What that common assumption fails to grasp is the similarity of mental labor to computation and the computational capacity of the human brain, which still outpaces the sum of all supercomputers on Earth. While user interface for the human brain is weak and requires years of conditioning, someone in a mental labor role is essentially performing many of the same tasks that computers would. The human brain is particularly good at identifying and correlating arbitrary data, which provide data-processing shortcuts to some solution.
A task type that forms a major dependency of AI-driven businesses is in the data itself: data science, data processing infrastructure and user interface. In this way, task types like these will be needed at ever-increasing rates, but eventually, software, data-science and infrastructure can all be automated. If someone lacks the equivalent of a master’s degree in either math or a field where math is strongly applied, they are unlikely to be useful for data-science.
Over the next two decades, older members of the workforce will be hit harder than younger members. This is because the demand for skills is shifting and our education’s capacity to imbue these skills is accelerating. Double down on your IRA and keep an eye on your pension fund.
All in all, the labor markets will favor those with highly varied skill sets, which results in a paradox. Whereas automation leaves many unemployed in an economy where high variety and proficiency of skills determine employability, UBI as a policy detracts from the incentive to work and to develop a highly varied set of skills. Why work if you don’t have to? Why learn or hone your skills in a difficult to understand, ever-changing economy where the potential payout for each acquired skill decreases year after year? What results is a cultural feedback loop that magnifies the effect that automation has on the set of behaviors humans reinforce. While automation and technology make us more capable as a society, in the long term, it may result in broadly disempowered individuals. Those with highly varied skill sets remain employable and work to create tools which render their own skills obselete.
What’s left is a socioeconomic structure that neither provides incentives to work nor rewards hard work proportionately, where the citizens must maintain technology they don’t understand built with skills they mostly lack. They lack these skills because, while they were in school, education didn’t offer children a significant perceived benefit, at least economically. In turn, these cultural trends could render the domestic economy vulnerable to global competition. Also, there are further interesting intergenerational dynamics that could eventually emerge, where the oldest generation has the widest variety of skills, which are mostly not propagated to subsequent generations. Many of these assumptions depend on the particular capabilities, practicality and limitations of neural interfaces.
And this all occurs during the same period of time when generalized AI may develop into a threat. When that time comes, shouldn’t humanity be as capable as possible? What we have to do is strive against this paradox and resist being lulled into complacency by utopia.
So we know that high unemployment rates are inevitable, given the trajectory of technology and its impact on labor markets. We also know that continued technological development is inevitable because any alternative implies unsustainable population growth and/or a transition of economic power to whoever develops the best technology. So the question we need to ask is: how do we deal with unemployment and other economic turbulence?
Policy Solutions Include:
★ Do nothing
★ Provide UBI (Ubiquitous Income)
★ National Service
Both of the latter generally involve growing the size of government and increasing its range of influence on our lives. Still, doing nothing is far worse. Both UBI and National Service are expensive result in wide-reaching second and third order socioeconomic effects. However, upon examining the potential higher order effects, one finds that National Service is far preferable. Many of the advantages of UBI, like opportunity for innovation and experimentation, can be realized through much smaller grant programs.
Generally, we find that many of these domestic issues and their root causes could be avoided or prevented outright, if only people were more educated, responsible, insightful or temperate. How do you catalyze your population to develop a robust set of knowledge and skills? To imbibe a set of beliefs that naturally causes them to live more responsibly? Given the right opportunities, the right incentives and an accurate understanding of the world, people generally make the best decisions for themselves and their families. Many of our domestic problems are a result of ignorance, lack of foresight, myopic desire and hopelessness.
The root causes of domestic issues in America are best addressed through cultural transformation. Without such a transformation, America is sure to succumb to domestic problems, like unrest, poverty, economic turbulence and unemployment.
National citizen service can be harnessed to imprint a near-universal experience or widely-shared types of experiences. If these experiences are common, they form the basis of pattern languages that help us communicate and diagnose domestic issues. UBI does not provide this opportunity for cultural transformation or the imprinting of types of experiences. Unless one assumes the best case to represent the majority (e.g. that people will react by altruistically using UBI as an opportunity for personal development instead of hedonism or netflix), then national service results in much preferable outcomes, socioculturally. IMO, this altruistic best case would be rare.
The development of connections in the social graph is driven through interpersonal experiences. Many domestic issues in America are caused by social isolation or at least could be ameliorated by methodically reducing isolation in the social graph. These include poverty, individual cases of mental heath, and the tendency for individuals to become radicalized. All of these problems are worse in the context of isolation.
In combination with data science, programs like national service can be leveraged to methodically weave the social graph into one that is stronger for all of us. When we have more connections, we have more options. When there are more connections, information flows around the social network. It converges faster and more completely. For example, this can inform us of problems with friends or family or opportunities for employment. This can weave together communities where individuals otherwise might feel isolated. This stronger social net encourages people to come together.
In the past, religion has been used to imprint experience types and to encourage social flux, which induces transmission of information and beliefs amoung its participants. Religious customs that are social, like regular attendance of church, drive social flux that triggers permeation of information through social networks. There are evolutionary dynamics to religions and belief systems. There are real benefits to their practices. The religions that managed to hold on to believers were able to convey and provide value.
Korea and Isreal are both countries which require mandatory military service for young adults. Such mandatory service can be leveraged to develop highly varied skill sets, making people more employable.
Mobilizing dozens of millions and integrating them all into a military force is not an option. And a military force wouldn’t exactly be welcomed for international development of infrastructure. We’ll need to expand a force like Americorps or create new institutions. Human capital to help build the foundations of underdeveloped economies, which provides a framework for economic growth in the 21st century.
So it’s 2036…
There’s an oversaturated labor market where low demand and high specificity in requirements for job roles renders several dozen million American adults unemployed. A higher proportion of the US government budget is dedicated to expansive entitlement programs with less left over for the discretionary budget.
What do you do?
You fucking hope you played your cards right because it’s going to be tough no matter what you do. Large amounts of unemployed people typically leads to unrest. It’s better for people to be busy or at least have their net human potential coordinated in some positive way. Throughout history, we’ve been able to depend on economic structures and self-interest to motivate people to cooperate and build. Yet, at this point, economics leads us to a place where that doesn’t naturally happen. Our incentivization structures break down, but we still need to remain competitive because the world hasn’t quite congealed into one.
Eden of the East - Opening Scene in Washington, DC
We leverage our human capital to posture ourselves for maximal growth
Whatever we do to address the economic turbulence, we need to develop a highly skilled labor force. If we’re going to have to spend the money, it should render our citizenry capable. So, to posture for growth, we can rebuild infrastructure in America and build out economic opportunities through cooperation with underdeveloped economies.
Furthermore, there is an upcoming soft-power showdown with China. Two decades of monopolar Western hegemony has yielded a naive generation of Americans. We need to focus significant effort on building ties and options for soft power and economic growth. China is mobilizing quickly on projects in Pakistan and Africa. In order for China to sustain GDP growth, they must expand. America, being multicultural, has innately greater capacity to build bridges with the rest of the world. In the long-term, these relationships are much more likely to be seen as authentic and reciprocal.
The operational and logistical overhead in utilizing the government as an employer of last resort sadly makes UBI seem more worthwhile. With UBI, you simply deposit money in bank accounts and maybe mail some checks. There’s no real operational overhead or complex logistics.
With national service, there is still significant disruption to the domestic economy. How do you train and mobilize this many people while minimizing disruption to private enterprises which can’t easily finance the labor?
Yet, if one is compelled to make a decision between UBI and an employer of last resort, the capacity to seize on the latter to hit fifty birds with one stone is undeniably more advantageous. It also reduces the risk of large swaths of population being either idle or feeling like their work and self-investment is rewarded disproportionately. The 21st century yields one of the great filters of Fermi’s paradox and we want to piss away all our money?
If we do nothing or choose UBI, we will see revolution and transformation into something requiring national service anyways. After any revolution, you have to placate the masses and continue to suppress dissent. How do you think that would happen?
There are again intergenerational and geoeconomic dynamics of the progression of unemployment via automation. These must be major considerations when architecting the operational and logistical aspects of a national service program. It’s important to understand that America isn’t the only country facing this socioeconomic transformation. The whole world must undergo this transition, but, like modernization and industrialization, it will not unfold simultaneously.
When the government is paying the living expenses of 15-50% of the population, it’s hard to imagine this being economically sustainable. Financing this requires that the government expend either taxes or credit. If there is correlation between tax volume and employment, which is certainly the case before the 21st century, then the government expenditures to mitigate unemployment will be unsustainable, if not at first, then eventually.
From an entrepreneur’s perspective, the purpose of automation is increase profit margins by reducing costs and improving efficiency. Yet, a long-term result of automation is unemployment that necessitates higher government expenditures, increasing the tax burden on corporations.
Taxing automation is one sensible policy option to placate economic turbulence, but this paradoxically makes these businesses less able to compete with those from nations that don’t tax automation. How did globalization affect domenstic manufacturing and why? Corporations moved overseas because they harnessed the cheapest source of labor available. That’s economics at work and similar dynamics will be in play if automation is taxed, though in some form it is a good idea.
If the policy response results in increased economic opportunity, while enabling American enterprises to remain competitive, then it will sustainable in the short-term. If it’s not sustainable in the long-term, then hopefully we got something good with all that money. Ultimately, that will be what we are left with.
For example, if we utilize National Service to rebuild America’s infrastructure, then we have infrastructure. If we utilize National Service to invest in underdeveloped economies, then American business can leverage those global markets as a framework for further growth. And, in case of robot apocalypse, our population has a highly-developed skill set as a result. But, if we write $30,000 annual checks so that 15-50% of our citizens can watch two-thirds of the Netflix catalog, then we’re left with a population who communicates poorly using sitcom references, but isn’t employable or able to contribute to society.
Economic growth ultimately stems from the net contributions of individuals. It’s built from the ground up. We know that we are going to have dozens of millions of unemployed between ages 20 and 60 by the 2030’s. The point is to invest that human potential and manpower to get the greatest return possible in anticipation of greater future threats to human existance. More challenging than the operational aspects is convincing Americans that this is not only necessary, but should be enthusiastically embraced when technology seems to remove most constraints. To do this we need a cultural transformation to produce a society built on values and character with a strong sense of identity and duty.
Governance in Space and the Post-Westphalian Era
The fourth part in this series is my fave. It covers the transition to space, the accompanying economic shifts and the evolution of governence mechanisms. As we approach the apex of globalization, we reach the end of the Westphalian era. We find that transnationalism wedges itself in the frame of international law and is used to contort precedent. Then, as we begin to colonize space, humanity again rearchitects our outermost paradigms of governance, leading to a transition away from the brief post-Westphalian period.